Padrões biogeográficos da distribuição dos anfíbios da Amazônia

Abstract

In recent years, rapid advancements in computing have revolutionized ecological research, enabling the development of advanced computational models to better understand ecosystems and biodiversity distribution. Using techniques such as artificial intelligence, machine learning, and complex simulations, these models analyze large volumes of data, predict ecological patterns and processes with high precision. In this context, the quality of primary biodiversity data is crucial for calibrating and validating these models, as well as supporting conservation policies and environmental management. However, there are significant biases and gaps in biodiversity knowledge, especially in remote areas like the Amazon, due to its vast geographic expanse, complex and diverse biological richness, and logistical and infrastructural challenges. Various sampling biases affect the representativeness of biodiversity studies, including taxonomic, spatial, and temporal biases, limiting the true understanding of biodiversity and ecological processes in the Amazon. In this Thesis, we constructed a robust database for 951 amphibian species in the Amazon, through which we highlighted patterns of richness and endemism among Amazonian amphibians, as well as biases and biogeographical gaps in their sampling. Our data demonstrated that the Amazon is the biome with the highest richness and endemism of amphibians, harboring 11% of the global amphibian richness with 82% endemism. Distribution patterns showed a longitudinal growth trend, with the western region of the Amazon exhibiting greater richness and endemism. However, amphibian sampling rates showed biases, with navigable rivers being the main influencing variable in sampling. Thus, we observed high biogeographical ignorance in amphibian sampling, particularly in the southern and southeastern regions of the Amazon. Recognizing and mitigating these biases and data gaps is crucial for understanding biodiversity. Improving sampling in tropical regions like the Amazon requires innovation in data collection techniques, international collaboration, and ongoing investment in research and scientific infrastructure. This is essential for a holistic and accurate understanding of Amazonian biodiversity, crucial to supporting effective conservation policies and environmental sustainability.